Mineralogical controls on Li, Sr and oxygen isotope composition of mixed CaMg carbonate phases with implications for sedimentary dolomites

The formation of ordered dolomite is unlikely to occur under ambient Earth's surface conditions, yet “disordered dolomite,” has been shown to crystallize at temperatures as low as 40 °C. Such synthetic precipitates have a similar d(104) spacing as dolomite, and have been studied previously to determine their O isotope compositions as a function of temperature with the goal of using oxygen isotopes as a temperature proxy. However, laboratory synthesis yields mineralogical assemblages that transform to more stable phase assemblages over time. Previous studies however have not thoroughly addressed how this transformation proceeds and how it affects O isotope compositions of the precipitates. To better understand the relationship between temperature and δ¹⁸O values of Ca-Mg‑carbonates at temperatures <100 °C, in this study, CaMg carbonates were synthesized at 40, 60 and 80 °C and incubated up to 104 days. Mineralogical composition was quantified using Rietveld refinement of X-ray diffraction patterns, while concomitantly monitoring fluid and solid compositions to assess the utility of the δ¹⁸O_solid, Li, and Sr compositions as paleo-proxies in complex Ca-Mg‑carbonate assemblages. The results suggest a continuous transformation of the mineralogy of the samples throughout the duration of the experimental runs, although Mg/Ca of the bulk solids remained quasi-constant at ∼1, and between 40 and 104 days of reaction the bulk δ¹⁸O_solid values did not exhibit significant variations. These δ¹⁸O_solid values were used to estimate temperature-dependent oxygen isotope fractionation between bulk solid and fluid that can be expressed as:${10}^3\ln {\alpha }_{\mathrm{solid}-\mathrm{fluid}}=1.78\left(\pm 0.13\right)\frac{{10}^6}{T^2}+8.47\left(\pm 1.20\right)$

where T is temperature in Kelvin.

The use of this relation yields significantly different temperature dependence compared to that reported earlier by Schmidt et al. (2005) using the same synthesis procedure. The difference can be assigned to mineralogical changes occurring in the precipitates over the course of the 104-day runs that may not have occurred in the earlier study owing to a shorter experimental duration. Here we discuss in detail the role mineralogy has on the chemical and isotopic compositions of CaMg carbonates, and the implications for using CaMg carbonate minerals as paleoarchives.